CN104076353A - Area target echo beam center velocity measurement method - Google Patents

Abstract

The invention discloses an area target echo beam center velocity measurement method. Firstly, echoes are subjected to acceleration compensation; secondly, whether a velocity filtering value is in a near-zone measurement pattern or a far-zone measurement pattern is judged, if the velocity filtering value is in the near-zone measurement pattern, after a frequency spectrum of the echoes is obtained through FFT operation, the Chirp-Z method is used for thinning spectral lines nearby the Doppler velocity to obtain an echo frequency spectrum, and if the velocity filtering value is in the far-zone measurement pattern, the echo frequency spectrum is obtained through the FFT operation, and finally the echo frequency spectrum is subjected to smoothing and envelop interception to obtain a frequency value corresponding to a beam center, so that the velocity value is calculated; finally, the Kalman filtering method is adopted to track the velocity, and the velocity, acceleration and other information at the next moment are predicted. By means of the area target echo beam center velocity measurement method, the requirements for a large dynamic range and high accuracy are met, and measurement accuracy is improved.

Description

A kind of Area Objects echo beam center speed measurement method

Technical field

The present invention relates to a kind of echo beam center speed measurement method, relate in particular to a kind of continuous wave Doppler radar high-precision surface target echo beam center speed measurement method, belong to Space Microwave remote sensing technology field.

Background technology

Continuous wave velocity radar is the crucial unit in moon survey of deep space landing engineering, and it provides accurate satellite velocities information for satellite guidance, navigation and control system, thereby ensures the safety of landing.The measuring accuracy of continuous wave velocity radar is higher, adopts single utilization high-precision surface target echo doppler beam center acquiring method, Area Objects Doppler frequency spectrum is calculated to the Doppler frequency of beam center.In addition, continuous wave velocity radar is all having broad application prospects aspect missile guidance and ship-board aircraft, unmanned plane landing.

The electromagnetic wave of antenna transmission is the wave beam with certain width, the radial component difference of speed on each component direction in wave beam, also with regard to corresponding different Doppler shifts, therefore the echoed signal that antenna is received is to be formed by stacking by multiple components with different frequency, because lander exists acceleration, can cause doppler bandwidth broadening, due to the impact of the factors such as scattering, can make the position of peak value component in echoed signal be offset simultaneously.

In order to improve the measuring accuracy of Doppler signal, must eliminate the impact of the Doppler frequency spectrum broadening that lander acceleration of motion causes, so need to carry out estimation and the compensation of acceleration, current existing document is all to the motion compensation under point target motion of automobile condition, can adopt maximum likelihood method, Time-Frequency Analysis Method, solution linear frequency modulation method estimating target acceleration based on acceleration template, the needed operand of these methods is all excessive, is unfavorable for real-time processing; The multinomial phase tranformation of second order (DPT) is the method for the estimated acceleration that a kind of operand is less, but for the acceleration estimation of Area Objects, its precision will reduce greatly.In addition, for the detection of Area Objects Doppler frequency, in existing document, adopt leading edge detection, energy barycenter method or geometrical center method, but these methods for asking for Doppler frequency corresponding to antenna beam center, evaluated error is poor.

Summary of the invention

The technical matters that the present invention solves is: overcome the deficiencies in the prior art, a kind of Area Objects echo beam center speed measurement method is provided, taken into account great dynamic range and high-precision requirement, improved measuring accuracy.

Technical solution of the present invention is: a kind of Area Objects echo beam center speed measurement method, and step is as follows:

(1) antenna reception to rf echo signal after microwave channel frequency conversion demodulation, obtain base-band analog signal, this base-band analog signal is converted to digital radar echoed signal s (n) through AD acquisition module;

(2) utilize Kalman filtering to previous measuring period Area Objects echo beam center speed carry out tracking filter, and the acceleration of this cycle Area Objects echo beam center is estimated, obtain the pie slice value v of previous cycle Area Objects echo beam center and the acceleration estimation value a of this cycle Area Objects echo beam center;

(3) utilize the acceleration estimation value a of this cycle Area Objects echo beam center that step (2) obtains, the digital radar echoed signal s (n) that step (1) is obtained carries out acceleration compensation;

(4) the pie slice value v of the previous cycle Area Objects echo beam center that determining step (2) obtains is in close-in measurement pattern or far field measurement pattern, if in close-in measurement pattern, and execution step (5); Otherwise execution step (6);

(5) the digital radar echoed signal after acceleration compensation is carried out to FFT analysis of spectrum and obtain digital radar echo Doppler frequency spectrum, digital radar echoed signal according to digital radar echo Doppler frequency spectrum after to acceleration compensation is carried out Chirp-Z conversion, obtain the digital radar echo Doppler frequency spectrum after frequency spectrum refinement, execution step (7);

(6) the digital radar echoed signal after acceleration compensation is carried out to FFT analysis of spectrum, obtain digital radar echo Doppler frequency spectrum, execution step (7);

(7) digital radar echo Doppler frequency spectrum is carried out to smoothing processing, the digital radar echo Doppler frequency spectrum after smoothing processing is carried out to envelope intercepting, calculate on this basis Area Objects echo beam center velocity amplitude;

(8) repeated execution of steps (1)-(7), obtain the real-time speed value of Area Objects echo beam center.

The implementation of described step (2) is:

If the third-order matrix of pie slice value is x_filter[], the third-order matrix of velocity estimation value is x_estimate[], velocity measurement is v _m, x_filter=x_estimate+K*z;

Wherein

z＝v _m-H*x_estimate

$K=\left(\frac{1}{H*P*H^T+R}\right)*P*H^T$

P=φ * P _{a upper cycle}* φ ^t+ Q

x_estimate＝φ*x_filter

P _{a upper cycle}for the error covariance matrix after upper one-period renewal;

After each computation of Period completes, upgrade according to the following formula error covariance matrix:

P=(I-K*H) * P, wherein I is unit matrix;

On the basis of the above, obtain pie slice value v=x_filter[0]; Acceleration estimation value a=x_estimate[1];

Wherein, H is observing matrix, $H=\left[\begin{array}{c}1\\ 0\\ 0\end{array}\right];$ φ is state-transition matrix, $\mathrm{\φ}=\left[\begin{array}{ccc}1& T& \frac{T^2}{2}\\ 0& 1& T\\ 0& 0& 1\end{array}\right];$ Q is dynamic noise covariance matrix, $Q=\left[\begin{array}{ccc}\frac{T^4}{4}& \frac{T^3}{2}& \frac{T^2}{2}\\ \frac{T^3}{2}& T^2& T\\ \frac{T^2}{2}& T& 1\end{array}\right]{\mathrm{\σ}}_w^2;$ P is error covariance matrix, $P=\left[\begin{array}{ccc}1& \frac{1}{T}& \frac{1}{T^2}\\ \frac{1}{T}& \frac{2}{T^2}& \frac{3}{T^3}\\ \frac{2}{T^2}& \frac{3}{T^3}& \frac{6}{T^4}\end{array}\right]q_v^2;$ R is observation noise covariance matrix initial value, t is cycle interval, for target velocity dynamic noise variance, for observation speed variance.

The implementation of described step (3) is:

Digital radar echoed signal after acceleration compensation is

$s^{\text{'}}\left(n\right)=s\left(n\right)\×\exp (-j\frac{4\mathrm{\πa}}{\mathrm{\λ}}{n}^2)=A_s\exp \left(j\frac{4\mathrm{\πv}}{\mathrm{\λ}}n\right)$

Wherein AS is digital echo signal amplitude, and λ is radar emission continuous wave wavelength, and n is discrete digital sample sequence number, $s\left(n\right)=A_s\exp (j\frac{4\mathrm{\πv}}{\mathrm{\λ}}n+j\frac{4\mathrm{\πa}}{\mathrm{\λ}}{n}^2).$

The implementation of described step (4) is:

When the pie slice value v ∈ [80m/s～500m/s] of current one-period beam center, being far field measurement pattern, when v ∈ [36m/s～80m/s], is close-in measurement pattern.

In described step (5), according to digital radar echo Doppler frequency spectrum, the digital radar echoed signal after to acceleration compensation is carried out Chirp-Z conversion, and the implementation that obtains the digital radar echo Doppler frequency spectrum after frequency spectrum refinement is:

If time domain sequences x (n) length in the digital radar echoed signal after acceleration compensation is N, utilize Chirp-Z mapping algorithm to obtain M point spectral sample value X (z in Z plane _k) mode be: $X\left(z_k\right)=W^{k^2/2}\underset{n=0}{\overset{N-1}{\mathrm{\Σ}}}y\left(n\right)h(k-n),0\≤k\≤M-1;$

Wherein

$y\left(n\right)=x\left(n\right)A^{-n}{W}^{n^2/2}$

$h(k-n)=W^{-{(k-n)}^2/2}$

A and W are the parameter under polar coordinates, A ₀and W ₀be 1; θ ₀for the phase angle of the initial sampling spot of frequency spectrum sequence x (n), obtain by the information extraction obtaining from FFT analysis of spectrum; for sample interval, M is number of sampling.

The implementation of described step (7) is:

The mode of digital radar echo Doppler frequency spectrum being carried out to smoothing processing is: establish frequency spectrum sequence f (n') N altogether ₁individual, n' is frequency spectrum sequence number, smoothly counts as 2L+1, and after smoothing processing, sequence F (n') is expressed as so:

$F\left(n^{\text{'}}\right)=\left\{\begin{array}{cc}f\left(n^{\text{'}}\right)& n^{\text{'}=\mathrm{0,1},...,L-1}\\ \frac{1}{2L+1}\underset{m=n^{\text{'}}-L}{\overset{m=n^{=\text{'}}+L}{\mathrm{\Σ}}}& n^{\text{'}}=L,...,N_1-L-1\\ f\left(n^{\text{'}}\right)& n^{\text{'}}=N_1-L,...,N_1-1\end{array}\right\}$

The mode of the digital radar echo Doppler frequency spectrum after smoothing processing being carried out to envelope intercepting is: peak value and frequency coordinate corresponding to this amplitude peak of on the digital radar echo Doppler frequency spectrum after smoothing processing, selecting this section of spectrum amplitude, and determine amplitude threshold according to the main lobe width of signal to noise ratio (S/N ratio) and antenna beam, on frequency spectrum, search for to both sides from frequency coordinate corresponding to amplitude peak, in the time that the spectrum amplitude searching is less than the thresholding of setting, obtain two frequency coordinate values, be respectively minimum Doppler frequency f _minwith maximum doppler frequency f _max;

The mode that calculates beam center frequency estimation is: wherein f _beamfor doppler centroid, θ represents antenna beam main lobe width corresponding angle; Thereby obtain beam center velocity amplitude be wherein λ is radar emission continuous wave wavelength

The present invention compared with prior art has following beneficial effect:

(1) the present invention is in order to improve velocity survey precision, adopts acceleration compensation technology, and Doppler's component that in echoed signal, acceleration causes is fallen in compensation, eliminated echo Doppler frequency spectrum and produce because of the existence of acceleration the situation of broadening; Use Kalman filtering that the estimated value of velocity survey filter value and acceleration is provided simultaneously;

(2) the present invention is in order to take into account great dynamic range and high-precision requirement, velocity survey region is divided into far field tupe and near region tupe, utilize different sampling rates to cover different Doppler frequency scopes, and in the higher near region tupe of accuracy requirement, adopt the method for Chirp-Z conversion, improve spectral resolution;

(3) while asking for Doppler center, first carry out smoothing processing, its advantage is to improve the precision that Doppler center is estimated, can overcome landform and change the discontinuous situation of Doppler frequency spectrum causing simultaneously; On Doppler's power spectrum, carry out top-down envelope intercept method, reduced the possibility of antenna sidelobe introducing strong scattering target.

Brief description of the drawings

Fig. 1 is Area Objects echo beam center velocity survey process flow diagram;

Fig. 2 is that Chirp-z converts the spiral sampling schematic diagram on z-plane;

Fig. 3 is Chirp-z mapping algorithm theory diagram;

Fig. 4 is that Area Objects echo beam center frequency is asked for schematic diagram.

Embodiment

As shown in Figure 1, the present invention proposes a kind of high-precision surface target echo beam center speed measurement method, step is as follows:

(1) antenna reception to rf echo signal after microwave channel frequency conversion demodulation, obtain base-band analog signal, be converted to digital radar echoed signal s (n) through AD acquisition module;

(2) utilize Kalman filtering to previous measuring period beam center speed carry out tracking filter, and the acceleration of this cycle beam center is estimated, obtain the pie slice value v of previous cycle beam center and the acceleration estimation value a of this cycle beam center;

(3) utilize the acceleration estimation value a of this cycle beam center that step (2) obtains, the digital radar echoed signal s (n) that step (1) is obtained carries out acceleration compensation;

In the time not considering acceleration, continuous wave Doppler echoed signal is:

s(n)＝A _Sexp(j2πf _dn)

$f_d=\frac{2v}{\mathrm{\λ}}$

In the time considering acceleration, continuous wave Doppler echoed signal is:

s(n)＝A _Sexp(j2πf _dn)

$f_d=\frac{2(v+\mathrm{an})}{\mathrm{\λ}}$

$s\left(n\right)=A_s\exp (j\frac{4\mathrm{\πv}}{\mathrm{\λ}}n+j\frac{4\mathrm{\πa}}{\mathrm{\λ}}{n}^2)$

So, consider to carry out acceleration compensation on the basis of equal acceleration movement model, need to be multiplied by the factor to echoed signal $\exp (-j\frac{4\mathrm{\πa}}{\mathrm{\λ}}{n}^2);$

Digital radar echoed signal after acceleration compensation

$s^{\text{'}}\left(n\right)=s\left(n\right)\×\exp (-j\frac{4\mathrm{\πa}}{\mathrm{\λ}}{n}^2)=A_s\exp \left(j\frac{4\mathrm{\πv}}{\mathrm{\λ}}n\right)$

Wherein A _sfor digital echo signal amplitude, λ is radar emission continuous wave wavelength, and n is discrete digital sample sequence number.

(4) the pie slice value v of the previous cycle beam center obtaining according to step (2) judges that this pie slice value is in close-in measurement pattern or far field measurement pattern, if in close-in measurement pattern, execution step (5); Otherwise execution step (6);

According to velocity survey scope and desired precision, velocity survey scope is divided into far field measurement pattern and close-in measurement pattern, far field measurement pattern velocity range 80～500m/s, measuring precision prescribed 1 σ is 0.33%V, close-in measurement mode speed scope is-36～80m/s, and the highest 1 σ of measuring accuracy is 0.05m/s;

According to upper one-period velocity measurement, next cycle measurement pattern is switched, the pie slice value v of current one-period beam center is greater than 500m/s or when be less than-36m/s, measurement pattern is constant; Calculate according to Ka frequency range 8mm wavelength, far field pattern Doppler spread is 10KHz～125KHz, and under the prerequisite that meets sampling thheorem, it is 600KHz that employing rate is set, and near region pattern Doppler spread is-4.5KHz～10KHz that it is 300KHz that sampling rate is set.

(5) the digital radar echoed signal after acceleration compensation is carried out to FFT analysis of spectrum, obtain digital radar echo Doppler frequency spectrum, digital radar echoed signal according to digital radar echo Doppler frequency spectrum after to acceleration compensation is carried out Chirp-Z conversion, obtain the digital radar echo Doppler frequency spectrum after frequency spectrum refinement, execution step (7);

If the length of the digital radar echo time domain sequences x (n) after acceleration compensation is N, Fourier transform (FFT) is write as following form:

$X\left(k^{\text{'}}\right)=\underset{n=0}{\overset{N-1}{\mathrm{\Σ}}}x\left(n\right)e^{-j2\mathrm{\πn}{k}^{\text{'}}/N},k^{\text{'}}\∈[0,N-1]$

If the length of time domain sequences x (n) is N, obtain M point spectral sample value X (z in Z plane _k), 0≤k≤M-1, its Chirp-Z mapping algorithm adopting is specific as follows.

If z _k=AW ^-k, 0≤k≤M-1

In formula, A and W are plural number, are expressed as by polar form

A in formula ₀and W ₀for real number, in the time of k=0, have

$z_0=A_0{e}^{j{\mathrm{\θ}}_0}$

As can be seen here, in above formula, A determines analysis of spectrum starting point z ₀position; W ₀value determine the trend of spiraling of analysis path, represent two angles between adjacent analysis site.If W ₀<1, along with k increases, analysis site z _kwith for step-length is to inner rotary; W ₀inwardly rotation when >1.As shown in Figure 2, be called spiral sampling.The special A that works as ₀=1, and W ₀within=1 o'clock, rotate along unit circle.

By z _kthe transform formula of substitution x (n) obtains

$\begin{array}{c}X\left(z_k\right)=\underset{n=0}{\overset{N-1}{\mathrm{\&Sigma;}}}x\left(n\right){\left[{\mathrm{AW}}^{-k}\right]}^{-n}\\ =\underset{n=0}{\overset{N-1}{\mathrm{\&Sigma;}}}x\left(n\right)A^{-n}{W}^{\mathrm{kn}},0\&le;k\&le;M-1\end{array}$

Utilize following relational expression

$\mathrm{nk}=\frac{1}{2}[n^2+k^2-{(k-n)}^2]$

Can obtain

$\begin{array}{c}X\left(z_k\right)=\underset{n=0}{\overset{N-1}{\mathrm{\&Sigma;}}}x\left(n\right)A^{-n}{W}^{[n^2+k^2-{(k-n)}^2]/2}\\ =W^{k^2/2}\underset{n=0}{\overset{N-1}{\mathrm{\&Sigma;}}}x\left(n\right)A^{-n}{W}^{n^2/2}{W}^{-{(k-n)}^2/2}\end{array}$

Order

$y\left(n\right)=x\left(n\right)A^{-n}{W}^{n^2/2}$

$h\left(n\right)=W^{-n^2/2}$

?

$X\left(z_k\right)=W^{k^2/2}\underset{n=0}{\overset{N-1}{\mathrm{\&Sigma;}}}y\left(n\right)h(k-n),0\&le;k\&le;M-1$

Above formula explanation, the M point analysis of spectrum of the sequence x (n) that length is N can obtain y (n) by taking advantage of in advance, then calculates the linear convolution of y (n) and h (n), is finally multiplied by three steps obtain.The theory diagram of this algorithm as shown in Figure 3.In Fig. 3, regard the unit impulse response of a digital filter as, its output V (n)=y (n) * h (n).

A due to what use in this programme ₀=1, and W ₀=1 situation, so we discuss analysis to this.θ ₀for the phase angle of initial sampling spot, the information obtaining by extraction FFT analysis of spectrum obtains; for sample interval, M is number of sampling, and the region of refinement is by parameter θ ₀, determine together with M, the radian scope of selected segmental arc is work as θ ₀=0, when M=N, Chirp-Z conversion is exactly DFT.θ ₀value can just can bear, but due to the symmetry of real sequence spectrum, general parameters θ in actual spectrum refinement ₀, m selects to make segmental arc be in [0, π] scope.X (z _k) digital angular frequency in the region of the corresponding refinement of (0≤k≤M-1) each spectral line is shown below:

Similar with DFT, if the sampling rate of sequence X (n) is f _s, its corresponding analog frequency is

From above formula, the frequency resolution of Chirp-Z conversion is less is M when larger, and frequency resolution is higher.

(6) the digital radar echoed signal after acceleration compensation is carried out to FFT analysis of spectrum, obtain digital radar echo Doppler region frequency spectrum;

If the length of the digital radar echo time domain sequences x (n) after acceleration compensation is N, Fourier transform (FFT) is write as following form:

$X\left(k^{\text{'}}\right)=\underset{n=0}{\overset{N-1}{\mathrm{\&Sigma;}}}x\left(n\right)e^{-j2\mathrm{\&pi;n}{k}^{\text{'}}/N},k^{\text{'}}\&Element;[0,N-1]$

(7) digital radar echo Doppler frequency spectrum is carried out to smoothing processing, the digital radar echo Doppler frequency spectrum after smoothing processing is carried out to envelope intercepting, calculate on this basis beam center velocity amplitude;

Taking wave beam R1 as example, Fig. 4 has provided R1 wave beam and has irradiated the schematic diagram of terrain object and echo frequency distribution thereof.Wherein, φ ₁for the angle of velocity reversal and horizontal direction, φ ₂for the angle of velocity reversal and R1 beam center.Wave beam orientation is θ to angle and the angle of pitch.F _min, f _maxminimum Doppler frequency and the maximum doppler frequency of corresponding echoed signal respectively, f _midfor the intermediate value of Doppler frequency, f _beamthe Doppler frequency value at corresponding beam center point place.

The minimum Doppler frequency f of echoed signal _minfor:

$f_{\min }=\frac{2V}{\mathrm{\&lambda;}}\cos ({\mathrm{\&phi;}}_2+\mathrm{\&theta;})$

The maximum doppler frequency f of echoed signal _maxfor:

$f_{\max }=\frac{2V}{\mathrm{\&lambda;}}\cos ({\mathrm{\&phi;}}_2-\mathrm{\&theta;})$

Can obtain the intermediate value f of Doppler frequency according to above formula _midfor:

$f_{\mathrm{mid}}=\frac{V}{\mathrm{\&lambda;}}\cos ({\mathrm{\&phi;}}_2+\mathrm{\&theta;})+\frac{V}{\mathrm{\&lambda;}}\cos ({\mathrm{\&phi;}}_2-\mathrm{\&theta;})$

The Doppler frequency value f at beam center point place _beamfor:

$f_{\mathrm{beam}}=\frac{2V}{\mathrm{\&lambda;}}\cos {\mathrm{\&phi;}}_2$

Can obtain thus:

$\frac{f_{\mathrm{mid}}}{f_{\mathrm{beam}}}=\frac{\cos ({\mathrm{\&phi;}}_2+\mathrm{\&theta;})+\cos ({\mathrm{\&phi;}}_2-\mathrm{\&theta;})}{2\cos {\mathrm{\&phi;}}_2}=\cos \mathrm{\&theta;}$

Arrangement can obtain:

$f_{\mathrm{beam}}=\frac{f_{\mathrm{mid}}}{\cos \mathrm{\&theta;}}$

Therefore the method that, we can take smoothing processing and envelope to intercept first obtains f _midvalue, then calculate according to above formula the frequency values that beam center is corresponding.

Smoothing processing is carried out on Doppler's power spectrum, on the one hand in order to improve Doppler center estimated accuracy, and the discontinuous situation of Doppler frequency spectrum of having avoided on the other hand topographic relief to cause.The mode of digital radar echo Doppler frequency spectrum being carried out to smoothing processing is: establish frequency spectrum sequence f (n') N altogether ₁individual, smoothly count as 2L+1, n' is frequency spectrum sequence number, after smoothing processing, sequence F (n') is expressed as so:

$F\left(n^{\text{'}}\right)=\left\{\begin{array}{cc}f\left(n^{\text{'}}\right)& n^{\text{'}=\mathrm{0,1},...,L-1}\\ \frac{1}{2L+1}\underset{m=n^{\text{'}}-L}{\overset{m=n^{=\text{'}}+L}{\mathrm{\&Sigma;}}}f\left(m\right)& n^{\text{'}}=L,...,N_1-L-1\\ f\left(n^{\text{'}}\right)& n^{\text{'}}=N_1-L,...,N_1-1\end{array}\right\}$

It is peak value from Doppler's power spectrum that envelope intercepts, top-down intercepting antenna beamwidth corresponding thresholding, and this method, has reduced antenna sidelobe and has introduced the situation of strong scattering target.The mode that envelope intercepts is: peak value and frequency coordinate corresponding to this amplitude peak of on the digital radar echo Doppler region frequency spectrum after smoothing processing, selecting this section of spectrum amplitude, and determine amplitude threshold according to the trunk width of channel ratio and antenna beam, on frequency spectrum, search for to both sides from frequency coordinate corresponding to amplitude peak, in the time that the spectrum amplitude searching is less than the thresholding of setting, obtain two frequency coordinate values, be respectively minimum Doppler frequency f _minwith maximum doppler frequency f _max;

The mode that calculates Estimation of Doppler central frequency value is:

Calculating beam center velocity amplitude is wherein λ is radar emission continuous wave wavelength.

(8) adopt kalman filter method to follow the tracks of the speed of previous cycle beam center, and the acceleration information of this cycle beam center is estimated, then repeated execution of steps (1)-(7), obtain the velocity amplitude of Area Objects echo beam center in real time.

Embodiment:

The digital radar echoed signal that definition AD gathers is s (n)=s_i (n)+j*s_q (n), the corresponding two kinds of tupes of sampling rate are respectively 600KHz and 300KHz, under two kinds of sampling rates, the signal integration time is respectively 25ms and 50ms, and it is 15000 points that digital signal samples is counted;

Treatment step of the present invention is as follows:

(1) antenna reception to rf echo signal after microwave channel frequency conversion demodulation, obtain base-band analog signal, be converted to digital radar echoed signal s (n)=s_i (n)+j*s_q (n) through AD acquisition module;

(2) utilize Kalman filtering to carry out tracking filter to the speed result of previous measuring period, and acceleration is estimated;

Observing matrix $H=\left[\begin{array}{c}1\\ 0\\ 0\end{array}\right]$

State-transition matrix $\mathrm{\&phi;}=\left[\begin{array}{ccc}1& T& \frac{T^2}{2}\\ 0& 1& T\\ 0& 0& 1\end{array}\right]$

Dynamic noise covariance matrix $Q=\left[\begin{array}{ccc}\frac{T^4}{4}& \frac{T^3}{2}& \frac{T^2}{2}\\ \frac{T^3}{2}& T^2& T\\ \frac{T^2}{2}& T& 1\end{array}\right]{\mathrm{\&sigma;}}_w^2$

Error covariance matrix $P=\left[\begin{array}{ccc}1& \frac{1}{T}& \frac{1}{T^2}\\ \frac{1}{T}& \frac{2}{T^2}& \frac{3}{T^3}\\ \frac{2}{T^2}& \frac{3}{T^3}& \frac{6}{T^4}\end{array}\right]q_v^2$

Observation noise covariance matrix initial value

Wherein T is that 128ms is got at cycle interval, for target velocity dynamic noise variance gets 144, for observation speed variance gets 9.If pie slice value third-order matrix x_filter[], velocity estimation value third-order matrix x_estimate[], velocity measurement is v _m;

First cycle: x_filter[0]=v _m;

Second period: x_filter[1]=(v _m-x_filter[0])/T;

x_filter[0]＝v _m；

The 3rd cycle: x_filter[2]=((v _m-x_filter[0])/T-x_filter[1])/T

x_filter[1]＝(v _m-x_filter[0])/T

x_filter[0]＝v _m；

x_estimate[]＝φ·x_filter[]；

After the 4th cycle:

x_filter＝x_estimate+K*z

Wherein

z＝v _m-H*x_estimate

$K=\left(\frac{1}{H*P*H^T+R}\right)*P*H^T$ Wherein P=φ * P _{a upper cycle}* φ ^t+ Q

P _{a upper cycle}for the error covariance matrix after upper one-period renewal

x_estimate＝φ*x_filter

Each cycle final updating error covariance matrix

P=(I-K*H) * P wherein I is unit matrix;

V=x_filter[0] be pie slice value

A=x_estimate[1] be the acceleration information that estimation obtains.

(3) the acceleration result obtaining according to step (2), the radar base-band digital echoed signal that step (1) is obtained is carried out acceleration compensation;

Acceleration compensation is to be multiplied by the factor to echoed signal the complex digital signal that obtains in step (1) gathering is s (n)=s_i (n)+j*s_q (n), with plural number the plural time domain sequences s'(n multiplying each other after being compensated);

(4) the pie slice value v of the previous cycle beam center obtaining according to step (2) judges that this pie slice value is in close-in measurement pattern or far field measurement pattern, if in close-in measurement pattern, execution step (5); Otherwise execution step (6);

(5) the digital radar echoed signal after acceleration compensation is carried out to FFT analysis of spectrum, obtain digital radar echo Doppler frequency spectrum X (k'), digital radar echoed signal according to digital radar echo Doppler frequency spectrum after to acceleration compensation is carried out Chirp-Z conversion, obtain the digital radar echo Doppler frequency spectrum after frequency spectrum refinement, then execution step (7);

To the time domain echoed signal after acceleration compensation, carry out FFT analysis of spectrum;

Plural time domain sequences s'(n after compensation) length be 15000, carry out N=16384 point Fourier transform (FFT) and write as following form:

$X\left(k^{\text{'}}\right)=\underset{n=0}{\overset{N-1}{\mathrm{\&Sigma;}}}x\left(n\right)e^{-j2\mathrm{\&pi;n}{k}^{\text{'}}/N},k^{\text{'}}\&Element;[0,N-1];$

Close-in measurement pattern, measuring accuracy requires high, adopts Chirp_Z conversion to carry out frequency spectrum refinement to echo Doppler region frequency spectrum; Utilize Chirp-Z to convert 512 frequency spectrums centered by spectrum peak coordinate as analysis spectrum, refinement multiple is 8 times, obtains 4096 frequency spectrum refinement results as beam center analysis spectrum X'(k') implementation be:

Plural time domain sequences s'(n after compensation) length be N=15000, utilize Chirp-Z mapping algorithm to obtain M=4096 point spectral sample value X (z in Z plane _k) mode be: $X\left(z_k\right)=W^{k^2/2}\underset{n=0}{\overset{N-1}{\mathrm{\&Sigma;}}}y\left(n\right)h(k-n),0\&le;k\&le;M-1,$ 0≤k≤M-1；

Wherein

$y\left(n\right)=s^{\text{'}}\left(n\right)A^{-n}{W}^{n^2/2}$

$h(k-n)=W^{-{(k-n)}^2/2}$

A ₀and W ₀be 1; θ ₀for frequency spectrum sequence s'(n) phase angle of initial sampling spot, for sample interval, M is number of sampling; X'(k')=X (z _k).

(6) the digital radar echoed signal after acceleration compensation is carried out to FFT analysis of spectrum, obtain digital radar echo Doppler frequency spectrum X (k'), 1024 frequency spectrums centered by peak coordinate in X (k') frequency spectrum are taken out as beam center analysis spectrum X'(k');

(7) to echo spectrum X'(k') carry out smoothing processing and envelope intercepting, thus carry out Doppler center estimated value;

To X'(k') beam center analysis spectrum carry out smoothing processing:

$X^{\text{'}\text{'}}\left(k^{\text{'}}\right)=\left\{\begin{array}{cc}{X}^{\text{'}}\left(k^{\text{'}}\right)& k^{\text{'}=\mathrm{0,1},...,L-1}\\ \frac{1}{2L+1}\underset{m=k^{\text{'}}-L}{\overset{m=k^{\text{'}}+L}{\mathrm{\&Sigma;}}}{X}^{\text{'}}\left(m\right)& k^{\text{'}}=L,...,N_1-L-1\\ X^{\text{'}}\left(k^{\text{'}}\right)& k^{\text{'}}=N_1-L,...,N_1-1\end{array}\right\}$

L gets 45;

First obtain amplitude peak A and the coordinate index of this section of frequency spectrum, establishing amplitude threshold is A/36, searches for to both sides from index position, in the time that the spectral magnitude searching is less than the thresholding of A/36, obtains minimum Doppler frequency f _minwith maximum doppler frequency f _max, establish θ=6 °, according to formula obtain doppler beam centre frequency f _beam, then according to formula obtain beam center velocity amplitude.

(8) adopt Kalman filtering method to follow the tracks of speed, and the acceleration information in next cycle is estimated, then repeated execution of steps (1)-(7), obtain the velocity amplitude of Area Objects echo beam center in real time.

Claims (6)

1. an Area Objects echo beam center speed measurement method, is characterized in that step is as follows:

(1) antenna reception to rf echo signal after microwave channel frequency conversion demodulation, obtain base-band analog signal, this base-band analog signal is converted to digital radar echoed signal s (n) through AD acquisition module;

(2) utilize Kalman filtering to previous measuring period Area Objects echo beam center speed carry out tracking filter, and the acceleration of this cycle Area Objects echo beam center is estimated, obtain the pie slice value v of previous cycle Area Objects echo beam center and the acceleration estimation value a of this cycle Area Objects echo beam center;

(3) utilize the acceleration estimation value a of this cycle Area Objects echo beam center that step (2) obtains, the digital radar echoed signal s (n) that step (1) is obtained carries out acceleration compensation;

(4) the pie slice value v of the previous cycle Area Objects echo beam center that determining step (2) obtains is in close-in measurement pattern or far field measurement pattern, if in close-in measurement pattern, and execution step (5); Otherwise execution step (6);

(5) the digital radar echoed signal after acceleration compensation is carried out to FFT analysis of spectrum and obtain digital radar echo Doppler frequency spectrum, digital radar echoed signal according to digital radar echo Doppler frequency spectrum after to acceleration compensation is carried out Chirp-Z conversion, obtain the digital radar echo Doppler frequency spectrum after frequency spectrum refinement, execution step (7);

(6) the digital radar echoed signal after acceleration compensation is carried out to FFT analysis of spectrum, obtain digital radar echo Doppler frequency spectrum, execution step (7);

(7) digital radar echo Doppler frequency spectrum is carried out to smoothing processing, the digital radar echo Doppler frequency spectrum after smoothing processing is carried out to envelope intercepting, calculate on this basis Area Objects echo beam center velocity amplitude;

(8) repeated execution of steps (1)-(7), obtain the real-time speed value of Area Objects echo beam center.

2. a kind of Area Objects echo beam center speed measurement method according to claim 1, is characterized in that: the implementation of described step (2) is:

If the third-order matrix of pie slice value is x_filter[], the third-order matrix of velocity estimation value is x_estimate[], velocity measurement is v _m, x_filter=x_estimate+K*z;

Wherein

z＝v _m-H*x_estimate

$K=\left(\frac{1}{H*P*H^T+R}\right)*P*H^T$

P=φ * P _{a upper cycle}* φ ^t+ Q

x_estimate＝φ*x_filter

P _{a upper cycle}for the error covariance matrix after upper one-period renewal;

After each computation of Period completes, upgrade according to the following formula error covariance matrix:

P=(I-K*H) * P, wherein I is unit matrix;

On the basis of the above, obtain pie slice value v=x_filter[0]; Acceleration estimation value a=x_estimate[1];

Wherein, H is observing matrix, $H=\left[\begin{array}{c}1\\ 0\\ 0\end{array}\right];$ φ is state-transition matrix, $\mathrm{\&phi;}=\left[\begin{array}{ccc}1& T& \frac{T^2}{2}\\ 0& 1& T\\ 0& 0& 1\end{array}\right];$ Q is dynamic noise covariance matrix, $Q=\left[\begin{array}{ccc}\frac{T^4}{4}& \frac{T^3}{2}& \frac{T^2}{2}\\ \frac{T^3}{2}& T^2& T\\ \frac{T^2}{2}& T& 1\end{array}\right]{\mathrm{\&sigma;}}_w^2;$ P is error covariance matrix, $P=\left[\begin{array}{ccc}1& \frac{1}{T}& \frac{1}{T^2}\\ \frac{1}{T}& \frac{2}{T^2}& \frac{3}{T^3}\\ \frac{2}{T^2}& \frac{3}{T^3}& \frac{6}{T^4}\end{array}\right]q_v^2;$ R is observation noise covariance matrix initial value, t is cycle interval, for target velocity dynamic noise variance, for observation speed variance.

3. a kind of Area Objects echo beam center speed measurement method according to claim 1, is characterized in that: the implementation of described step (3) is:

Digital radar echoed signal after acceleration compensation is

$s^{\text{'}}\left(n\right)=s\left(n\right)\&times;\exp (-j\frac{4\mathrm{\&pi;a}}{\mathrm{\&lambda;}}{n}^2)=A_s\exp \left(j\frac{4\mathrm{\&pi;v}}{\mathrm{\&lambda;}}n\right)$

Wherein A _sfor digital echo signal amplitude, λ is radar emission continuous wave wavelength, and n is discrete digital sample sequence number, $s\left(n\right)=A_s\exp (j\frac{4\mathrm{\&pi;v}}{\mathrm{\&lambda;}}n+j\frac{4\mathrm{\&pi;a}}{\mathrm{\&lambda;}}{n}^2).$

4. a kind of Area Objects echo beam center speed measurement method according to claim 1, is characterized in that: the implementation of described step (4) is:

When the pie slice value v ∈ [80m/s～500m/s] of current one-period beam center, being far field measurement pattern, when v ∈ [36m/s～80m/s], is close-in measurement pattern.

5. a kind of Area Objects echo beam center speed measurement method according to claim 1, it is characterized in that: in described step (5), according to digital radar echo Doppler frequency spectrum, the digital radar echoed signal after to acceleration compensation is carried out Chirp-Z conversion, and the implementation that obtains the digital radar echo Doppler frequency spectrum after frequency spectrum refinement is:

If time domain sequences x (n) length in the digital radar echoed signal after acceleration compensation is N, utilize Chirp-Z mapping algorithm to obtain M point spectral sample value X (z in Z plane _k) mode be: $X\left(z_k\right)=W^{k^2/2}\underset{n=0}{\overset{N-1}{\mathrm{\&Sigma;}}}y\left(n\right)h(k-n),0\&le;k\&le;M-1;$

Wherein

$y\left(n\right)=x\left(n\right)A^{-n}{W}^{n^2/2}$

$h(k-n)=W^{-{(k-n)}^2/2}$

A and W are the parameter under polar coordinates, A ₀and W ₀be 1; θ ₀for the phase angle of the initial sampling spot of frequency spectrum sequence x (n), obtain by the information extraction obtaining from FFT analysis of spectrum; for sample interval, M is number of sampling.

6. a kind of Area Objects echo beam center speed measurement method according to claim 1, is characterized in that: the implementation of described step (7) is:

The mode of digital radar echo Doppler frequency spectrum being carried out to smoothing processing is: establish frequency spectrum sequence f (n') N altogether ₁individual, n' is frequency spectrum sequence number, smoothly counts as 2L+1, and after smoothing processing, sequence F (n') is expressed as so:

$F\left(n^{\text{'}}\right)=\left\{\begin{array}{cc}f\left(n^{\text{'}}\right)& n^{\text{'}=\mathrm{0,1},...,L-1}\\ \frac{1}{2L+1}\underset{m=n^{\text{'}}-L}{\overset{m=n^{=\text{'}}+L}{\mathrm{\&Sigma;}}}& n^{\text{'}}=L,...,N_1-L-1\\ f\left(n^{\text{'}}\right)& n^{\text{'}}=N_1-L,...,N_1-1\end{array}\right\}$

The mode of the digital radar echo Doppler frequency spectrum after smoothing processing being carried out to envelope intercepting is: peak value and frequency coordinate corresponding to this amplitude peak of on the digital radar echo Doppler frequency spectrum after smoothing processing, selecting this section of spectrum amplitude, and determine amplitude threshold according to the main lobe width of signal to noise ratio (S/N ratio) and antenna beam, on frequency spectrum, search for to both sides from frequency coordinate corresponding to amplitude peak, in the time that the spectrum amplitude searching is less than the thresholding of setting, obtain two frequency coordinate values, be respectively minimum Doppler frequency f _minwith maximum doppler frequency f _max;

The mode that calculates beam center frequency estimation is: wherein f _beamfor doppler centroid, θ represents antenna beam main lobe width corresponding angle; Thereby obtain beam center velocity amplitude be wherein λ is radar emission continuous wave wavelength.